Sounds from intense auroral displays have been reported from ancient times. In his book Germania Cornelius Tacitus (56AD-120AD), the Roman historian, wrote of people from the northern part of Germany who claimed to hear them. The sounds are usually described as a faint swishing, hissing, sighing or rustling which rises and falls in time with the surges in brightness of the auroral display.
Most of the puzzling aspects of auroral sounds parallel the similar problems encountered with anomalous sounds from bolides, which is why a definitive theory remains to be securely established. The late Professor C A Chant, of the University of Toronto, was interested in the phenomenon throughout his life. He could not find an explanation which fitted all the facts, concluding that an electrical effect known as a brush discharge came closest to the truth (refer my bibliography, item 6).
Professor Chant's theory was enthusiastically embraced by two American researchers, S M Silverman and T F Tuan, who in 1973 wrote a comprehensive 110-page paper on the subject. It is a mine of information. The brush discharge theory is supported to some extent by D.E. Olsen, who measured a jump in the geoelectric field from the fair-weather value of around 100 volt per metre to over 10,000 volt per metre during an intense aurora.This is approaching the electric field strength needed to excite an audible brush discharge under suitable conditions, but there are two problems with this.
Witnesses who report hearing auroral sounds do not usually mention "St Elmo's Fire" effects accompanying them, which a brush discharge would produce. But more seriously, the brush discharge theory fails to account for the puzzling fact that a great many auroral observers fail to hear the sounds while others in the vicinity may hear them clearly.
These two problems are completely overcome by Keay's geophysical electrophonics explanation. Whether an observer perceives any sounds is dependent upon the presence of a nearby electric to acoustic transducer, accounting for the apparent capriciousness of the effect. Also it is known that auroras produce radiation in the ELF/VLF region of the electromagnetic spectrum, which corresponds to much of the audible range of the acoustic spectrum. When ELF/VLF signals are played back through an audio amplifier the output sounds bear a close resemblance to those reported by observers who have heard the sounds from an aurora.
There have been a number of attempts by scientists to record auroral sounds, without success until recently. The Director of an Arctic scientific institute once told me there were no sounds because his staff had often used the finest microphones and audio equipment to record them, without success. They had set the microphones out on the snow well away from trees and other objects which "might affect the result". It is precisely their presence which provides the necessary transduction of electrical to acoustic energy. Our anechoic chamber experiments (refer my bibliography, item 7) suggest that even the simple expedient of laying a piece of plain typing paper on top of the microphone could have made all the difference!
From a Swedish web-site comes the news that auroral sounds have at last been recorded: "The first successful attempt to record auroral sounds has been done. Eigil Ungstrup, scientist from Denmark, has recorded sounds by (the) help of a large radio antenna used in the(ir) ionosphere investigations. During a period when the antenna was not in use, he put a microphone in the focus of the antenna and recorded sounds from the northern lights. He describes the sounds as hissing..."
In this interesting experiment it seems most likely that parts of the antenna structure were acting as the transducing device and the sound intensity may have been enhanced by the sensible placement of the microphone.
The spectacle of a brilliant auroral displayis something never to be forgotten, especially for those fortunate enough to hear the sounds on the rare occasions when they occur.